Determine whether intravitreal injection of bevacizumab exerts long-term effects on neurodevelopmental outcomes in children with retinopathy of prematurity (ROP) when reaching the age of 8 years. In total, 277 children were enrolled. Patients were stratified into full-term, preterm without ROP, ROP without treatment, or ROP with treatment groups, based on gestational age and ROP status. Children under the gestational age of 37 weeks were considered premature. Patients' cognitive outcomes were evaluated using a full-scale intelligence quotient (full score and percentile) generated by the Wechsler Intelligence Scale for Children-Fourth Edition every 1 year to 2 years. At the mean age of 7.8 years, ROP without and with treatment groups demonstrated lower full-scale intelligence quotient scores and percentiles, compared with full-term and premature groups (both P < 0.05). Full-scale intelligence quotient scores and percentiles did not significantly differ between patients who received different treatments for ROP (full score P = 0.19; percentile P = 0.37). After adjusting for gestational age, logMAR best-corrected visual acuity was negatively associated with full-scale intelligence quotient scores ( P = 0.0008) and percentiles ( P = 0.0002). At the mean age of 8 years, patients with ROP undergoing injection of bevacizumab did not exhibit worse cognitive outcomes than those who underwent laser photocoagulation or both treatments. Gestational age and best-corrected visual acuity correlated with cognitive development in children.